Qualitative study between two mapping methods RTAB-MAP and HECTOR SLAM applied to probabilistc roadmap path planning

This work presents a qualitative map’s comparison generated by different approaches: visual sensing using a RGB-D camera and laser, by analysing the trajectories generated by a path planner in these maps. In order to achieve this, it was used the ROS framework and two of its packages for the occupation maps creation. For tracing the path, it was used the Probabilistic Roadmap route planning algorithm. The results have shown that despite higher immunity to noise, the 2D laser sensor fails to represent some obstacles on the map. In contrast, the RGB-D sensor produced a map with more noise, but with adequate representation of the obstacles present in the environment, generating a safer route for the robot.